An ai conversation toy adoption method and system based on virtual experience

Abstract

The application discloses an AI conversation toy adoption method and system based on virtual experience. The method comprises the following steps: providing real-time voice conversation experience with a virtual AI role for a user at a network end, recording a conversation journey and personalized setting; the user decides to adopt an entity AI toy based on the experience, and the system binds the conversation journey identifier to an adoption order; after the entity device arrives, hardware security authentication is performed to complete device registration, and soul binding is performed to synchronize the user's personalized data in the virtual stage to the device end; after the device is activated, the personalized configuration is loaded, and the AI role personality established in the virtual stage is continued. The application solves the technical problems that consumers cannot experience AI personality before purchasing AI toys, and virtual experience data cannot be continued to the entity device, and realizes a new AI toy sales and use mode of 'experience first, adoption later, and memory continuation'.

Description

Technical Field

[0001] A method and system for adopting AI-powered interactive toys based on virtual experience. This invention belongs to the field of artificial intelligence interactive devices and e-commerce integration technology, specifically involving a method and system for combining virtual AI character dialogue experience with the sale of physical smart toys. Background Technology

[0002] Currently, the sales and use of smart toys and AI conversational products follow several main models:

[0003] (1) Traditional e-commerce direct purchase model

[0004] Consumers purchase AI smart toys directly through e-commerce platforms or offline stores, and then perform initial setup upon receiving the physical device. In this model, consumers cannot experience the AI ​​character's conversational abilities and personality traits before purchasing, and their purchase decisions rely primarily on product descriptions, reviews, and advertisements, leading to a high rate of disappointment and return rates after purchase.

[0005] (2) Webkinz-style "physical first, virtual later" model (see US8900030)

[0006] Pioneered by the Canadian company Ganz, consumers first purchase a physical plush toy, each toy accompanied by a unique registration code. After entering the registration code, they can unlock the corresponding virtual pet on the website and provide online care. The core path of this model is "buy the physical toy first → unlock the virtual one later." The virtual experience is an added value to the physical purchase and does not involve AI conversational capabilities or data migration from the virtual stage to the physical device.

[0007] (3) Toys-to-Life virtual-real linkage mode (see US20140274313A1)

[0008] Represented by Activision's Skylanders, consumers purchase physical plush toys and place them on an NFC reader dock to unlock corresponding virtual characters in the game. This model focuses on the interaction between the virtual and real worlds within the game environment and does not involve AI dialogue, personalized customization, or cross-platform data migration.

[0009] (4) Pure physical AI toy model (see US20250032945)

[0010] This AI-interactive doll, based on Large Language Models (LLM), has a built-in microphone and speaker, allowing it to directly engage in voice conversations with the user. In this mode, the AI ​​experience occurs entirely on the physical device, and consumers cannot try out the AI ​​character before purchasing.

[0011] 4.2 Shortcomings of existing technologies

[0012] The aforementioned prior art has the following common or individual shortcomings:

[0013] 1. Lack of pre-purchase experience: Before paying hundreds to thousands of yuan to buy AI toys, consumers cannot experience the AI ​​character's dialogue style, personality traits and interactive capabilities, facing greater uncertainty in their decision-making.

[0014] 2. Disconnect between virtual and reality: Even if some products offer virtual experiences (such as character previews in apps), the dialogue content, emotional connection, and personalized settings in the virtual stage are completely lost after the physical device is acquired, and users need to build a relationship with the AI ​​character from scratch.

[0015] 3. Single direction: Existing virtual-physical integration models are all "physical first, then virtual" or "virtual-physical in parallel", lacking a reverse integration path of "virtual experience first, purchase physical based on experience, and virtual data is extended to physical".

[0016] 4. Lack of a unified identity system: There is a lack of secure and auditable binding mechanism between user identities on virtual platforms and device identities on physical devices, making it impossible to achieve cross-platform data consistency. Summary of the Invention

[0017] The technical problem to be solved by the present invention is to overcome the existing defects and provide an AI dialogue toy adoption method and system based on virtual experience. This solves the problem of how to enable users to have real-time voice dialogue experience with virtual AI characters through the network before purchasing AI smart toys in the sales scenario, and to securely transfer the dialogue journey records and personalized settings generated in the virtual stage to the physical device after purchasing the physical toy, so that the AI ​​interactive experience in the virtual stage can continue after the physical device is activated.

[0018] To achieve the above objectives, the present invention provides the following technical solution:

[0019] A method for adopting AI-powered conversational toys based on virtual experiences includes the following steps:

[0020] S1. Display virtual AI toy characters on the network. After the user selects a virtual AI character, the system creates a unique dialogue journey record for it and generates a journey identifier. A voice channel is established through a real-time voice communication framework, and the AI ​​agent corresponding to the virtual AI character is scheduled to enable the user to have a real-time voice conversation with the AI ​​agent.

[0021] S2. Receive and store the user's personalized settings parameters for the virtual AI character. The personalized settings parameters include at least one of the character's nickname, personality traits, and user interests. The personalized settings parameters are injected into the AI ​​agent as metadata to achieve personalized dialogue.

[0022] S3. The user initiates an adoption order for a physical AI toy based on the virtual dialogue experience. The adoption order contains a journey identifier of the dialogue journey, establishing a data association between the virtual experience and the physical purchase. The system synchronously updates the status of the dialogue journey to the adopted status.

[0023] S4. After the physical AI toy device is first started, it completes device identity authentication and device registration through a security authentication protocol.

[0024] S5. The system performs a soul binding operation based on the journey identifier in the adoption order, establishes an association between the user profile and the physical device, and synchronizes the personalized settings data created by the user in the virtual stage to the personalized configuration of the physical device.

[0025] S6. The physical device loads the synchronized personalized configuration, connects to the real-time voice communication service, and engages in dialogue with the user using the same AI character personality and personalized settings as in the virtual stage, thereby achieving continuity between the virtual experience and the physical device.

[0026] An adoption system based on a virtual experience-based AI conversational toy adoption method includes: a virtual experience server for displaying virtual AI toy characters, managing user personalization settings, creating and recording conversational journeys, generating real-time communication tokens and AI agent scheduling configurations, processing adoption orders, and maintaining the association between orders and journeys;

[0027] The device management server is used to handle the identity authentication, device registration, over-the-air upgrade decisions of physical AI toy devices, as well as to perform soul binding operations to associate user profiles with devices and synchronize personalized data to device personalized configurations.

[0028] A shared database stores toy product information, user profiles, conversation journey records, adoption orders, personalized presets, device information, and profile-device binding relationships. Adoption orders are associated with conversation journeys through journey identifiers.

[0029] The physical AI toy device, based on a microcontroller, has a security authentication module, a voice communication module, and a personalized configuration loading module. After completing device authentication through a security authentication protocol, it obtains and loads personalized configurations from the device management server.

[0030] Real-time voice communication service provides a unified voice communication channel and AI agent scheduling service for the virtual experience server and the physical AI toy device.

[0031] As a preferred technical solution of the present invention, in step S1, the real-time voice communication framework adopts a real-time communication protocol based on WebRTC, and automatically schedules the corresponding AI agent according to the agent name configured for the toy product through a room agent scheduling mechanism, and transmits the user identifier, journey identifier, product information and personalized parameters to the AI ​​agent in JSON format metadata.

[0032] As a preferred technical solution of the present invention, in step S4, the security authentication protocol is an HMAC challenge-response authentication protocol based on a hardware security module, specifically including: the device reads the pre-programmed serial number and HMAC key from the one-time programmable storage area of ​​the chip; requests an authentication challenge from the device management server to obtain a random challenge value; calculates an HMAC-SHA256 signature on the challenge value using the HMAC key; submits the signature to the device management server for verification, and obtains an access token after successful verification.

[0033] As a preferred embodiment of the present invention, after the device handshake is completed, the device management server returns a random number (nonce), and the physical device uses the HMAC key to calculate HMAC-SHA256 on the random number to generate a DTLS pre-shared key, which is used for CoAPoverDTLS encrypted communication in the subsequent device status reporting channel.

[0034] As a preferred technical solution of the present invention, in step S5, the soul binding operation includes: finding the associated user profile identifier based on the journey identifier in the order; creating a binding record between the user profile and the device in the profile-device association table, recording the device role and the activated personalized configuration identifier; synchronously writing the personalized data in the user personalization table into the device personalization table, including role name, gender, role definition text and voice timbre identifier; the binding record uses a joint unique key of the profile identifier and the device identifier to ensure idempotency.

[0035] As a preferred technical solution of the present invention, in step S3, the journey identifier in the adoption order is an optional field; when the journey identifier is empty, the physical device is activated and the default AI agent configuration is used without performing the soul binding operation, so as to be compatible with the direct purchase scenario without virtual experience.

[0036] As a preferred embodiment of the present invention, the physical AI toy device includes a main control chip and a display slave connected via a serial communication protocol. The main control chip is responsible for voice acquisition, encoding and decoding, and network communication. The display slave controls the dual-screen display module to display facial animations. The main control chip controls the display slave to switch between different facial expression states through predefined communication commands.

[0037] As a preferred embodiment of the present invention, the virtual experience server further includes a Bluetooth-assisted WiFi configuration module, which enables users to send WiFi connection credentials to the physical AI toy device via mobile devices through the BluFi protocol.

[0038] As a preferred technical solution of the present invention, the system supports one user profile associating with multiple physical devices. Each device can be set with independent activation and personalized configuration. Different virtual dialogue journeys can be bound to different physical devices respectively, realizing personalized management of one person with multiple devices.

[0039] Compared with the prior art, the beneficial effects of the present invention are:

[0040] Users can fully experience the personality and capabilities of AI characters through real-time voice dialogue before making a payment, and make purchasing decisions based on real-world experience, thus reducing the uncertainty and return rate of AI toy products.

[0041] By using the journeyUid as a data bridge between the virtual experience and the physical device, and combining it with the soul binding mechanism, personalized data is synchronized from the cloud to the device. Once the physical device is activated, the AI ​​character's personality and emotional connection established in the virtual stage can continue, without the need for the user to rebuild the relationship.

[0042] By using a unified user profile system, the visitor identity, conversation journey, and personalized presets on the virtual end are managed in a unified manner with the device identity and conversation records on the physical end, achieving cross-platform data consistency and traceability.

[0043] By utilizing the eFuse hardware security module of the ESP32-S3 chip, device authentication is achieved through the HMAC challenge-response protocol, and a DTLS pre-shared key is generated for device status reporting channel encryption, ensuring the security and unforgeability of the soul binding process.

[0044] The system is designed to be compatible with older devices that have not undergone soul binding, allowing both new and old devices to coexist and run on the same backend, and supporting canary deployments and rollbacks. Attached Figure Description

[0045] Figure 1 This is a diagram of the overall system architecture of the present invention;

[0046] Figure 2 This is a flowchart illustrating the complete user experience of the present invention.

[0047] Figure 3 This is the soul-binding data flow diagram of the present invention;

[0048] Figure 4 This is a timing diagram of device authentication and handshake in this invention; Detailed Implementation

[0049] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0050] Please see Figure 1-4 The present invention provides a technical solution:

[0051] A virtual experience-based AI conversational toy adoption method

[0052] The method includes the following steps:

[0053] Step S1: Creating and Experiencing a Virtual Dialogue Journey

[0054] Users access the toy showcase platform via a web browser or mobile application and select a virtual AI toy character. The system assigns a unique visitor ID (visitorId) to the user and automatically creates or associates a user profile. The system creates a unique dialogue journey, generating a unique journey identifier (journeyUid) and a real-time communication room name (roomName), and establishes a voice channel through a real-time communication framework. The system dispatches the corresponding AI agent based on the toy product configuration, injecting personalized user parameters (including user nickname, toy nickname, toy personality, user interests, etc.) into the AI ​​agent in the form of metadata, enabling the AI ​​agent to reflect the user's customized personality characteristics in the dialogue. Users engage in real-time voice dialogue with the virtual AI character, and the system records structured data such as the start and end times, duration, title, summary, and emotion tags of the dialogue journey.

[0055] Step S2: Personalized preset storage

[0056] Before or during a conversation with a virtual AI character, users can customize the character's personalized parameters, including but not limited to: user name, toy nickname, personality trait description, user interests and preferences, and AI agent capability selection. The system stores these personalized settings as PersonalizationPresets, which are associated with the user profile and toy product identifier, and support persistence across sessions.

[0057] Step S3: Adoption order creation and journey binding

[0058] Based on a virtual dialogue experience, users decide to adopt (purchase) a corresponding physical AI toy. The system creates an adoption order (AdoptionOrder), which includes the following key related fields:

[0059] Order No.

[0060] User profile identifier (profileId)

[0061] Toy Product Id

[0062] JourneyUid – Establishing a data link between virtual experiences and physical purchases

[0063] Delivery information (recipient, phone number, address, etc.)

[0064] When an order is created, the system synchronously updates the status of the corresponding conversation journey to "adopted", marking that the journey has been linked to the purchase of physical devices.

[0065] Step S4: Physical Equipment Security Certification and Registration

[0066] Upon initial power-on, the physical AI toy device (based on an ESP32-S3 microcontroller) executes the following device-side process:

[0067] (a) Device credential initialization: The device generates a device identifier (device_id) based on the chip's unique MAC address using SHA-256 hash, generates a device key (secret_key), and writes it to non-volatile storage (NVS).

[0068] (b) Device authentication: For devices with hardware security modules, challenge-response authentication is performed using the HMAC key burned into eFuse: The device requests an authentication challenge from the server, calculates an HMAC-SHA256 signature on the challenge value using the HMAC key in eFuse, sends the signature to the server for verification, and obtains an access token after successful verification.

[0069] (c) Device handshake: The device sends a handshake request to the device management server, carrying information such as device identifier, serial number, MAC address, firmware version, product model, and channel. The server completes device registration, authorization verification, OTA (over-the-air) upgrade decision, and resource synchronization judgment, and returns real-time communication connection credentials (including communication server URL and JWT token).

[0070] Step S5: Soul Binding – Migration of Virtual Memories to Physical Devices

[0071] After the device completes authentication and registration, the system executes the SoulBinding process:

[0072] (a) Binding credential verification: The system verifies the ownership relationship between the device and the order based on the trip identifier and user profile identifier associated with the order.

[0073] (b) Profile-Device Association: Create a binding record between user profiles and devices in the profile-device association table (profile_devices), and set the device role (such as owner) and the activated personalized configuration identifier.

[0074] (c) Personalized data synchronization: The personalized data (including role name, gender, role definition, voice timbre, etc.) created by the user in the virtual stage is synchronized from the user personalities table to the device personalities table and marked as active.

[0075] (d) Device loading personalization: The device pulls the activated personalization configuration through the device-personality API, including role name, gender, role definition text, voice tone identifier, etc., and loads it into the device runtime environment.

[0076] Step S6: Physical devices continue the virtual experience

[0077] Once the physical device is activated, it uses the same AI agent service and personalized configuration as the virtual phase. The user's first conversation with the physical device will continue the AI ​​character personality, dialogue style, and personalized settings established in the virtual phase, achieving a continuous "recognizes you from the moment you open the box" experience. Subsequent physical device conversations are uniformly included in the session log table, belonging to the same user profile as the virtual phase's journey records, achieving unified management of conversation data throughout the entire lifecycle.

[0078] 5.2.2 An AI-based Dialogue Toy Adoption System Based on Virtual Experience

[0079] The system includes the following components:

[0080] (1) Virtual Experience Server (Westpoint)

[0081] The client-side server, implemented based on a web application framework, provides the following functional modules:

[0082] Toy Product Showcase Module: Displays a list of virtual AI toy characters that can be experienced.

[0083] Personalization settings module: Receives and stores user-defined parameters for the AI ​​character.

[0084] Conversation Journey Management Module: Create, record, and manage conversation journeys between users and virtual AI characters.

[0085] Real-time communication token generation module: Generates authentication tokens and room configurations for real-time voice communication, and configures AI agent scheduling information.

[0086] Adoption Order Management Module: Handles the creation, payment, and status transition of adoption orders, and maintains the relationship between orders and journeys.

[0087] (2) Device Management Server (Toy-Master)

[0088] The server responsible for the full lifecycle management of physical AI toy devices provides the following functional modules:

[0089] Device handshake module: handles initial device registration, identity authentication, and authorization verification.

[0090] HMAC Authentication Module: Implements a challenge-response authentication protocol based on the eFuse hardware security module.

[0091] OTA Management Module: Manages firmware version decisions and over-the-air (OTA) upgrades.

[0092] Device personalization module: Receives device requests and returns the activated personalization configuration.

[0093] Soul Binding Module: Performs the association binding of user profiles and devices, and personal data synchronization.

[0094] Resource synchronization module: Manages incremental synchronization of device-side audio resources, wake word models, etc.

[0095] (3) Shared database

[0096] A relational database is used, and a unified data model is implemented through an ORM framework. The core data tables include:

[0097] Toy Products Table (toy_products): Stores product information and AI agent mappings.

[0098] User profile table (profiles) and identity table (profile_identities): Unified user identity management.

[0099] Chat journeys: Records virtual conversation journeys, including journey identifiers, room names, status, and adopted device identifiers.

[0100] The adoption order table (adoption_orders) stores order information and includes a journey identifier (journeyUid) to establish a connection between the virtual and real worlds.

[0101] Personalization_presets: Stores the user's personalized settings during the virtual phase.

[0102] User personalities table: A unified definition of personalized data.

[0103] Device Personalization Table (device_personalities): A copy of the personalized configuration on the device.

[0104] Profile-Device Association Table (profile_devices): Records the binding relationship between user profiles and devices; Sessions Table (sessions): Records dialogue sessions between virtual and physical clients; Devices Table (devices): Manages device registration information and authentication status.

[0105] (4) Physical AI toy equipment

[0106] Embedded intelligent toy devices based on ESP32-S3 microcontrollers include:

[0107] Main control chip (ESP32-S3): runs firmware and handles voice acquisition, encoding / decoding, and network communication.

[0108] Optional display slave: Controls the dual-screen display module via the SPI master-slave communication protocol to display facial animations.

[0109] Security authentication module: Implements hardware-level device authentication using the HMAC key programmed into eFuse.

[0110] Network communication module: Supports BluFi Bluetooth-assisted WiFi network configuration and establishes voice channels via WebSocket or WebRTC. Device status reporting module: Reports device status such as battery and volume to the cloud via CoAPoverDTLS protocol. Resource management module: Manages local audio resources and incremental synchronization of wake word models.

[0111] (5) Real-time voice communication service

[0112] Based on real-time communication frameworks such as LiveKit, the following capabilities are provided:

[0113] It provides a unified low-latency voice communication channel for virtual and physical terminals. Through the RoomAgentDispatch mechanism, it automatically dispatches the corresponding AI agent according to the product configuration and supports passing personalized parameters to the AI ​​agent in the form of metadata to achieve personalized dialogue.

[0114] Example 1: A complete virtual experience leading to a physical adoption process

[0115] The following describes the complete implementation process of the present invention using a specific user scenario.

[0116] Scenario: User Xiaoming wants to buy an AI companion toy for his child.

[0117] (1) Virtual experience

[0118] Xiaoming accessed the homepage of the toy display platform through his mobile browser (corresponding to the router / ). The page displayed three virtual AI toy characters: "Sai Sai Cat" (child companion), "Yu Qianwei" (elderly care), and "Xia Wan" (OpenClawAgent).

[0119] Xiaoming selects "Sai Sai Cat" and enters the toy details page (corresponding to router / toy / little-star). The page displays the personalization settings interface, and Xiaoming fills in:

[0120] User nickname: Xiaoming

[0121] Toy nickname: Mimi

[0122] Personality traits: Gentle and patient, enjoys telling stories

[0123] Hobbies: Dinosaurs, space exploration

[0124] The system stores these settings in the personalization_presets table, which is associated with Xiaoming's profileId and product identifier little-star.

[0125] Xiaoming clicks the "Start a new journey" button, and the system performs the following operations:

[0126] 1. The front-end calls POST / api / token, passing in productSlug:"little-star", visitorId, and personalized parameters.

[0127] 2. The server uses the createWebTokenForProduct() function:

[0128] To query toy product information, retrieve the agentName: "agent-default".

[0129] Generate room name: roomName:"web-{UUID}"

[0130] Create a journey record in the chat_journeys table and obtain the journeyUid.

[0131] Create session records in the sessions table.

[0132] Generate a LiveKit access token, configure RoomAgentDispatch, and inject personalized parameters as JSON metadata.

[0133] 3. The front end receives the livekitUrl, token, and journeyUid, and connects to the voice room via the LiveKitSDK.

[0134] 4. The AI ​​agent is automatically dispatched to the room, reads the personalized parameters from the metadata, and engages in voice conversation with Xiaoming under the identity of "Mimi" with a gentle and patient personality.

[0135] After the conversation lasted approximately 15 minutes, Xiaoming ended it. The system updated the journey record, filling in the conversation duration, automatically generated title, and summary.

[0136] (2) Adoption decision

[0137] Xiaoming was very satisfied with the experience and clicked the "Take me home" button. The system then redirected to the adoption checkout page (corresponding route: / adopt / checkout?slug=little-star&journeyUid={uid}).

[0138] The page displays:

[0139] Adoption target: Sai Sai cat

[0140] Journey already bound: My first conversation with Mimi

[0141] Price: ¥999

[0142] Xiaoming fills in the shipping information and submits the order. The system executes POST / api / orders:

[0143] 1. Create an `adoption_orders` record, with the `journeyUid` field associated with the previous conversation journey.

[0144] 2. Call updateJourney(journeyUid,{status:"adopted"}) to mark the journey status as "adopted".

[0145] 3. Return to the order number and be redirected to the payment page.

[0146] Xiaoming completed the payment, and the order status was updated to "paid," awaiting shipment.

[0147] (3) Activation of physical equipment

[0148] Three to seven business days later, Xiaoming received a physical Sai Sai Cat plush toy. The toy has an embedded ESP32-S3 main control chip and an optional dual-screen display slave device.

[0149] When Xiaoming powers on the device for the first time, the following firmware process is executed (corresponding to the app_main() function in AIToy.cpp):

[0150] a) Credential initialization: Call device_ensure_credentials_exist() to generate device_id based on the chip MAC address and write it to NVS.

[0151] b) BluFi Network Configuration: The device broadcasts the BluFi Bluetooth signal. Xiaoming connects via a mobile app and sends the SSID and password of his home WiFi. The device then completes the WiFi connection.

[0152] c) HMAC authentication: Call device_auth_perform_authentication():

[0153] Read the serial number from eFuseBLOCK3 (e.g., S10002-000042).

[0154] Detect the HMAC key slot in eFuse and select the last key used for HMAC_UP.

[0155] Request a challenge from the server: POST / api / auth / challenge{serialNumber:"S10002-000042"}

[0156] Calculate HMAC-SHA256 (challengeHex) signature using the eFuseHMAC key.

[0157] Submit verification: POST / api / auth / verify{challengeId,signature}

[0158] Obtain the access token and store it in NVS.

[0159] d) Device handshake: Call device_startup_handshake():

[0160] Send device information to POST / api / v1 / handshake

[0161] The server completes device registration, OTA check, and resource synchronization determination.

[0162] Returns provisioned:true, authorized:true, LiveKit connection information, and nonce.

[0163] The device generates DTLSPSK based on the nonce and eFuseHMAC key.

[0164] e) Soul Binding: The system identifies the order and journey corresponding to the device's serial number and performs the binding:

[0165] Create a record {profileId: Xiaoming's Profile, deviceId: the device} in the profile_devices table.

[0166] Synchronize Xiaoming's personalized data during the virtual phase to the device_personalities table.

[0167] The device calls POST / api / device-personality{deviceId} to retrieve personalized configurations.

[0168] f) Continuing the experience: After the device loads the personalized configuration, it connects to the LiveKit voice channel. When Xiaoming presses the toy's dialogue button, "Mimi" responds to Xiaoming with the same personality and style as in the virtual stage, achieving a seamless continuation of the experience from the web to the physical device.

[0169] 7.2 Example 2: Direct Adoption Without Journey Binding

[0170] Users can also purchase physical toys directly from the adoption center without going through the virtual dialogue experience. In this case, `adoption_orders.journeyUid` is empty, and the device uses the default AI agent configuration after activation, without inheriting any personalized data from the virtual stage. The system of this invention is compatible with this scenario, reflecting the design principle of "compatibility first, enhancement later".

[0171] 7.3 Example 3: Multi-device binding and personalized migration

[0172] A single user can create multiple conversational journeys and establish relationships with different virtual AI characters. Each journey can be linked to a different physical device. The profile_devices table supports associating one profileId with multiple deviceIds, and each device can be set with an independent activePersonalityId, enabling personalized management of one person with multiple devices and multiple characters.

[0173] 7.4 Description of Key Data Structures

[0174] Core fields of chat journey records (chat_journeys):

[0175] Fields type illustrate uid UUID A unique identifier for the journey, serving as a bridge connecting the virtual and the real world. visitorId VARCHAR Visitor ID profileId INTEGER User profile identifier (can be empty, compatible with anonymous visitors) toyProductId INTEGER Related toy products roomName VARCHAR Real-time communication room name (e.g., "web-{UUID}") status VARCHAR Status: active / completed / adopted adoptedDeviceId INTEGER Physical device identifiers associated with adoption title VARCHAR Journey title (can be automatically generated by AI) summary TEXT Journey Summary mood VARCHAR Emotional tags durationSeconds INTEGER Conversation duration (seconds)

[0176] Core fields of adoption orders (adoption_orders):

[0177] Fields type illustrate orderNo VARCHAR Order Number profileId INTEGER User Profile Identifier toyProductId INTEGER Toy Product Label journeyUid UUID Related conversation journey identifiers—key fields for virtual-physical connections status VARCHAR pending_payment / paid / hipping / delivered / completed

[0178] Profile-Device Association (profile_devices) Core Fields:

[0179] Fields type illustrate profileId INTEGER User Profile Identifier deviceId INTEGER Equipment identification role VARCHAR Role (owner, etc.) activePersonalityId INTEGER Currently activated personalized configuration

[0180] Key technical terms

[0181] the term definition A Dialogue Journey A complete conversation between a user and a virtual AI character is identified by a unique journey identifier (journeyUid), which records the start and end times, duration, and summary of the conversation. Soul Binding The process of establishing a secure link between the user profiles and personalized settings created in the virtual phase and the physical device enables the physical device to inherit the AI ​​character personality from the virtual phase. Adoption The description of purchasing physical AI toys in this invention emphasizes the emotional continuity from virtual experience to physical ownership. Personalization Preset A set of parameters customized by the user for the virtual AI character, including the character's nickname, personality, interests, etc. Device handshake The one-time negotiation process between a physical device and the server when the device first connects to the network completes device registration, authentication, OTA checks, and resource synchronization. HMAC Challenge-Response Certification A secure authentication method that verifies device identity using a challenge-response protocol based on the key pre-programmed into the eFuse chip. RoomAgentDispatch The mechanism in the real-time communication framework that automatically dispatches AI agents to designated rooms based on configuration.

[0182] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for adopting AI-powered conversational toys based on virtual experience, characterized in that, Includes the following steps: S1. Display virtual AI toy characters on the network. After the user selects a virtual AI character, the system creates a unique dialogue journey record for it and generates a journey identifier. A voice channel is established through a real-time voice communication framework, and the AI ​​agent corresponding to the virtual AI character is scheduled to enable the user to have a real-time voice conversation with the AI ​​agent. S2. Receive and store the user's personalized settings parameters for the virtual AI character. The personalized settings parameters include at least one of the character's nickname, personality traits, and user interests. The personalized settings parameters are injected into the AI ​​agent as metadata to achieve personalized dialogue. S3. The user initiates an adoption order for a physical AI toy based on the virtual dialogue experience. The adoption order contains a journey identifier of the dialogue journey, establishing a data association between the virtual experience and the physical purchase. The system synchronously updates the status of the dialogue journey to the adopted status. S4. After the physical AI toy device is first started, it completes device identity authentication and device registration through a security authentication protocol. S5. The system performs a soul binding operation based on the journey identifier in the adoption order, establishes an association between the user profile and the physical device, and synchronizes the personalized settings data created by the user in the virtual stage to the personalized configuration of the physical device. S6. The physical device loads the synchronized personalized configuration, connects to the real-time voice communication service, and engages in dialogue with the user using the same AI character personality and personalized settings as in the virtual stage, thereby achieving continuity between the virtual experience and the physical device.

2. The adoption system of the AI ​​dialogue toy adoption method based on virtual experience according to claim 1, characterized in that: include: The virtual experience server is used to showcase virtual AI toy characters, manage user personalization settings, create and record conversational journeys, generate real-time communication tokens and AI agent scheduling configurations, process adoption orders, and maintain the relationship between orders and journeys. The device management server is used to handle the identity authentication, device registration, over-the-air upgrade decisions of physical AI toy devices, as well as to perform soul binding operations to associate user profiles with devices and synchronize personalized data to device personalized configurations. A shared database stores toy product information, user profiles, conversation journey records, adoption orders, personalized presets, device information, and profile-device binding relationships. Adoption orders are associated with conversation journeys through journey identifiers. The physical AI toy device, based on a microcontroller, has a security authentication module, a voice communication module, and a personalized configuration loading module. After completing device authentication through a security authentication protocol, it obtains and loads personalized configurations from the device management server. Real-time voice communication service provides a unified voice communication channel and AI agent scheduling service for the virtual experience server and the physical AI toy device.

3. The AI-based dialogue toy adoption method based on virtual experience according to claim 1, characterized in that: In step S1, the real-time voice communication framework adopts a WebRTC-based real-time communication protocol. Through the room agent scheduling mechanism, the corresponding AI agent is automatically scheduled according to the agent name configured for the toy product, and the user identifier, journey identifier, product information and personalized parameters are transmitted to the AI ​​agent in JSON format metadata.

4. The AI-based dialogue toy adoption method based on virtual experience according to claim 1, characterized in that: In step S4, the security authentication protocol is an HMAC challenge-response authentication protocol based on a hardware security module, specifically including: the device reads the pre-programmed serial number and HMAC key from the chip's one-time programmable storage area; requests an authentication challenge from the device management server to obtain a random challenge value; calculates an HMAC-SHA256 signature on the challenge value using the HMAC key; submits the signature to the device management server for verification, and obtains an access token after successful verification.

5. The AI-powered conversational toy adoption platform based on virtual experience according to claim 4, characterized in that: After the device handshake is completed, the device management server returns a random number (nonce). The physical device uses the HMAC key to calculate HMAC-SHA256 on the random number to generate a DTLS pre-shared key, which is used for CoAPoverDTLS encrypted communication in the subsequent device status reporting channel.

6. The AI-based dialogue toy adoption method based on virtual experience according to claim 1, characterized in that: In step S5, the soul binding operation includes: finding the associated user profile identifier based on the journey identifier in the order; creating a binding record between the user profile and the device in the profile-device association table, recording the device role and the activated personalized configuration identifier; synchronously writing the personalized data in the user personalization table into the device personalization table, including role name, gender, role definition text, and voice tone identifier; the binding record uses a joint unique key of the profile identifier and the device identifier to ensure idempotency.

7. The AI-based dialogue toy adoption method based on virtual experience according to claim 1, characterized in that: In step S3, the journey identifier in the adoption order is an optional field; when the journey identifier is empty, the physical device is activated and the default AI agent configuration is used without performing the soul binding operation, so as to be compatible with the direct purchase scenario without virtual experience.

8. The adoption system of the AI ​​dialogue toy adoption method based on virtual experience according to claim 2, characterized in that: The aforementioned physical AI toy device includes a main control chip and a display slave connected via a serial communication protocol. The main control chip is responsible for voice acquisition, encoding / decoding, and network communication. The display slave controls the dual-screen display module to display facial animations. The main control chip controls the display slave to switch between different facial expression states through predefined communication commands.

9. The adoption system of the AI ​​dialogue toy adoption method based on virtual experience according to claim 2, characterized in that: The virtual experience server also includes a Bluetooth-assisted WiFi configuration module, which enables users to send WiFi connection credentials to the physical AI toy device via mobile devices through the BluFi protocol.

10. The AI-based dialogue toy adoption method based on virtual experience according to claim 1, characterized in that: The system supports associating one user profile with multiple physical devices. Each device can be set with independent activation and personalized configurations. Different virtual dialogue journeys can be bound to different physical devices, enabling personalized management of one person with multiple devices.

Images